List your company

Test-Time Scaling Selection on RouterBench Cost-Sensitive

0.8337RewardOracle
0.5437480.6190240.69430.769576May 29, 2026

Evaluation results

MethodLinks
OracleNumber of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.833757.3810.6
OracleBackbone=Qwen3-0.6B, Number of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.829754.8710.5
UNISCALEBackbone=Qwen3-0.6B, Number of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.753546.523.3
GreedyBackbone=Qwen3-0.6B, Number of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.718443.7554
UNISCALENumber of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.707946.8849.4
TSBackbone=Qwen3-0.6B, Number of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.702244.567.9
MLPBackbone=Qwen3-0.6B, Number of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.700641.7548.6
NeuralUCBBackbone=Qwen3-0.6B, Number of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.69844270.6
k-NNBackbone=Qwen3-0.6B, Number of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.696636.549
RandomBackbone=Qwen3-0.6B, Number of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.693742.1276.7
k-NNNumber of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.65941.38326
MLPNumber of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.630147644
NeuralUCBNumber of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.58850.751,558.2
RandomNumber of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.5731531,358.7
GreedyNumber of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.558945660.6
TSNumber of seeds=5, Iterations=160, Warm-up phase=50-step2026.050.554952.121,762.8